{"id":"W2909244445","doi":"10.1145/3278721.3278728","title":"Towards Provably Moral AI Agents in Bottom-up Learning Frameworks","year":2018,"lang":"en","type":"article","venue":"","topic":"Adversarial Robustness in Machine Learning","field":"Computer Science","cited_by":18,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo","funders":"","keywords":"Computer science; Top-down and bottom-up design; Artificial intelligence; Cognitive science; Programming language; Psychology","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0006501717,0.0002029913,0.0002241648,0.0002044549,0.0002065103,0.0002581323,0.001185007,0.0002855532,0.0003073983],"category_scores_gemma":[0.0005131169,0.000188178,0.00006004724,0.0008105446,0.0001070418,0.0008664584,0.0007913662,0.001810593,0.0002979212],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001014753,"about_ca_system_score_gemma":0.0001295679,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0003399812,"about_ca_topic_score_gemma":0.00004158199,"domain_scores_codex":[0.9979488,0.0002032081,0.0002993394,0.0005732633,0.000429704,0.0005456686],"domain_scores_gemma":[0.9990973,0.00006773444,0.000100239,0.0005063093,0.0001191708,0.0001091766],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00007632977,0.0002008009,0.1911886,0.0000466401,0.00006002944,0.0001840609,0.0120147,0.05353571,0.0002032018,0.2812607,0.007254147,0.453975],"study_design_scores_gemma":[0.0005808501,0.0001729633,0.02254648,0.00005200268,0.000003610009,0.00001054328,0.000101551,0.9629583,0.0003409872,0.008046394,0.004840341,0.0003459513],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.03346322,0.00001366011,0.9421284,0.002392979,0.001268843,0.0001690451,1.126655e-7,0.0004301328,0.02013362],"genre_scores_gemma":[0.846245,0.000001941308,0.1491509,0.001557379,0.0002871362,0.00001055865,0.000001326007,0.0000186568,0.002727148],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.9094226,"threshold_uncertainty_score":0.7866226,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02601395670154006,"score_gpt":0.3166067087152982,"score_spread":0.2905927520137582,"validation_status":"score_only:v0-immature-baseline","note":"Baseline scores from an immature model (maturity gate not passed). Scores rank; they never assert a category."}}